The Detection Of The Embezzled Signal In Satellite Communications

The embezzlement detection is a new subject in satellite communication field. Using the satellite transponder's transparency, the illegal organization occupies the idle frequency band, or adds DS signals to the normal operation signals. Both affect the normal operation. For the former, the classic theory of signal detection and estimation methods can be used to solve the problem; for the latter, it is very difficult to be detected because the embezzled DS signals submerge in the background noise. So the further research of DS signal embezzlement detection is very important to the security of satellite communications.This thesis describes the basic principle and characteristic of satellite communication and DSSS(direct sequence spread spectrum) communication system, and analyzes the signal-simplified model of satellite transponder. The signal detection and parameter estimation are studied for DSSS/BPSK signal with low signal-to-noise ratio, including power spectrum method, time-domain auto-correlation method, and cyclic spectrum method. The simulation of three methods validates the detection capability. According to the characteristic of satellite communication and DSSS communication, the detection methods based on cyclic auto-correlation characteristic, and fourth-order cumulant characteristic are put forward. In the former method, we need weaken the energy of satellite signal firstly, and then detect the cyclic auto-correlation characteristic. The efficiency is enhanced through searching the max of cyclic auto-correlation matrix in this method. In the latter one, we directly detect the embezzled signal according to the fourth-order cumulant 2D slice characteristics, without pretreatment to the received signals.The results of simulation show that both the methods could detect the embezzled signal in the mixed signals in a higher signal to interference ratio.